If your portfolio holds ten different cryptocurrencies, it may still be a single bet. September 21, 2026 demonstrates the point: of the 20 non-stablecoins among the 25 largest cryptocurrencies, all 20 closed higher that day, 15 of them by at least five percent in 24 hours. Anyone who believed they were broadly diversified noticed nothing on this Monday, because everything rose at once. On a day with the opposite sign, that same lockstep becomes very noticeable indeed.
So we measured how closely the large cryptocurrencies actually move together: twelve coins, 66 pairs, 90 trading days. The detailed result is further down. The short version: average pairwise correlation stands at 0.61, and over the past 30 days it is higher than the quarterly average.
What correlation means for cryptocurrencies
Correlation is a metric between minus 1 and plus 1 that describes how evenly two prices move together. At plus 1, two coins move in perfect lockstep; at 0, their daily moves have nothing to do with one another; at minus 1, they move in exactly opposite directions. It is not calculated on the prices themselves but on daily returns, meaning the percentage changes from day to day.
For your portfolio, that is the decisive quantity behind the word diversification. Diversification does not mean holding many names, it means holding positions that do not fall at the same time. Two coins with a correlation of 0.90 are, in a downturn, effectively the same position with two tickers. The technical term for the consequence is concentration risk: on paper the portfolio looks spread out, and when it matters it behaves like a single holding.
The time axis matters. Correlations are not constants. They typically rise in phases where a shared driver moves the whole market, such as an interest rate decision or a liquidity shock, and they fall when individual coins follow news of their own. That is precisely why a single snapshot is worth little and a comparison of two time windows is informative.
15 of 20 top coins above five percent: the finding from September 21
Bitcoin traded at $85,369 or €74,333 on September 21, 2026 at 12:59 UTC, putting it 6.1 percent above the previous day. The total market came to $2.907 trillion, up 3.42 percent, and bitcoin dominance stood at 58.8 percent. All figures are calls to the CoinGecko programming interface from that same moment.
What is remarkable about the day is not the size of the move but its breadth. These 15 of the 20 non-stablecoins among the 25 largest cryptocurrencies were up at least five percent over 24 hours: Bitcoin, Ethereum, XRP, Solana, Zcash, Dogecoin, Monero, WhiteBIT Coin, Rain, Chainlink, Cardano, Stellar, NEAR Protocol, Bitcoin Cash and Avalanche. Not one of the 20 was down. A day like that is normal in a market-wide move, and it is the occasion to stop taking the diversification in your own portfolio as given.
Method: 66 coin pairs across 90 trading days
This analysis was carried out by cryptoticker.io on September 21, 2026. The method in one sentence: for twelve cryptocurrencies, daily closing prices in US dollars were retrieved via the CoinGecko programming interface, logarithmic daily returns were derived from them, and the Pearson correlation coefficient was calculated for each of the 66 possible pairs.
The measurement period covers the 90 daily returns from June 22 to September 20, 2026. The current day is left out because it is not yet complete. Objects examined: twelve coins, namely Bitcoin, Ethereum, BNB, XRP, Solana, Dogecoin, Cardano, Chainlink, Avalanche, Stellar, Monero and Zcash, and the 66 pairs formed from them. Stablecoins are left out because their prices are pegged to a currency and a correlation calculation says nothing there.
Ethereum, XRP and Solana: where correlation to bitcoin sits above 0.77
Measured across the 90 daily returns through September 20, 2026, correlation to bitcoin looks like this:
- Ethereum: 0.89
- XRP: 0.87
- Solana: 0.81
- Dogecoin: 0.81
- Chainlink: 0.78
- Cardano: 0.77
- BNB: 0.72
- Stellar: 0.72
- Zcash: 0.61
- Avalanche: 0.57
- Monero: 0.33
Six of the eleven coins measured sit at 0.77 or above. A portfolio built from Bitcoin, Ethereum, XRP, Solana, Chainlink and Cardano therefore spreads capital across six names whose daily moves are three quarters explained by the same market move. Across all 66 pairs the average is 0.61; 38 pairs reach 0.60 or more, and eleven pairs come in at 0.80 or more. The highest pair is Bitcoin and Ethereum at 0.89.
Monero, Avalanche and Zcash: the three outliers in the measurement
Three readings fall outside the pattern. Monero comes in at 0.33, making it the only one of the twelve coins measured whose daily moves are mostly not explained by the broader market. Avalanche follows at 0.57, Zcash at 0.61. The lowest pair in the entire measurement is Avalanche and Monero at 0.03, meaning virtually no relationship across 90 days.
For Monero, an obvious explanation lies in its own news flow: the coin is under regulatory pressure, several trading venues have removed it from their offering in the European Union, and events like that move the price regardless of what bitcoin is doing. That is an attribution, not a measured cause. The measurement itself says only that the lockstep is weaker, not why.
A low reading is incidentally not a verdict on quality, and certainly not a buy recommendation. Monero traded at $570.63 or €496.87 on September 21, 2026, 7.47 percent above the previous day, so it too rose with the market that day. Low correlation merely means a coin follows risks of its own. Those own risks can be larger than the market risk you are diluting with them.
Correlation is rising: 0.61 over 90 days, 0.64 over 30 days
Run the same measurement over only the last 30 daily returns and the average across all 66 pairs rises from 0.61 to 0.64. The lockstep has therefore increased, and for individual coins markedly so: Ethereum climbs from 0.89 to 0.93, Cardano from 0.77 to 0.90, XRP from 0.87 to 0.90. Monero also ticks up slightly, from 0.33 to 0.37, but remains the outlier.
In practical terms: the protection that diversification within the crypto market offers was weaker over the past month than in the quarter before it. Anyone who set their portfolio allocation in early summer and has not looked at it since is working with outdated assumptions. That is no reason to rush, but it is a reason to run the numbers again.
Concentration risk in your portfolio: how to check your own diversification
The figures above apply to twelve large coins, not to your portfolio. Your allocation, your weightings and your time period produce different values. You can work it out without statistics software, and it pays to do so in this order.
Start with the weightings. Write down what share of total value each position represents. Very often the answer is already there: if bitcoin and ether together make up 70 percent and eight further coins fill the remaining 30 percent, the correlation of the small positions decides almost nothing. A portfolio tracker takes over this part and displays the weightings continuously, which saves doing the arithmetic by hand.
The second step is to look at shared drivers beyond the price statistics. Note which of your positions run on the same blockchain, are held in custody by the same trading platform or depend on the same stablecoin. Those connections appear in no correlation figure and still take effect simultaneously when it matters. An outage at the custodian platform hits every position held there at once, no matter how differently their prices behave in normal times.
The third step is a stress test by hand. Take the sharpest single-day loss in the total market over the past twelve months, apply that same percentage to every position and work out the portfolio value. That overstates the damage for positions with low correlation, and for exactly that reason it gives you the lower bound at which you can tell whether the size of your overall position fits your situation.
Leverage and liquidation price: why lockstep shortens the distance
For leveraged positions, high correlation is more than a theoretical problem. Anyone running several positions on shared collateral, meaning in the same margin account, sees every position move into the red at once during a market-wide move. The collateral then melts away not proportionally but across all positions simultaneously, and the liquidation price approaches faster than the calculation for a single position would suggest.
The point to check is therefore not the distance to liquidation per position but the distance for the account as a whole. Assume that all your positions make the same percentage move, because at correlations around 0.90 that is the realistic assumption. Two further quantities belong on the same sheet of paper: the funding rate, meaning the ongoing payment between the long and short side on perpetual contracts, and the stop you set in advance.
Holding period under section 23 of the German Income Tax Act: what applies when you reallocate
Anyone who wants to reallocate after this exercise should look at the tax first, because it changes the outcome. For private investors in Germany, cryptocurrencies count as other assets within the meaning of section 23 of the German Income Tax Act. After a holding period of more than one year a disposal gain is tax free; within the year it is taxable as a private disposal transaction, and an exemption threshold of €1,000 per calendar year applies to the sum of those gains.
The decisive point when reallocating is that swapping one coin for another is a disposal for tax purposes. Swap cardano into monero to lower your correlation and you have disposed of cardano and started a fresh one-year clock for monero. If the cardano position is still under a year old, the gain becomes taxable even though not a single euro has landed in your bank account. A reallocation made for risk reasons can be sensible and still trigger a tax charge that exceeds the benefit.
On the place of custody, the provider’s licence is worth a look. Which companies are permitted to provide crypto asset services in Germany is listed by BaFin in its register of crypto institutions. That is a question of authorisation and says nothing about price risk, but it belongs among the shared drivers from the second step above.
What this measurement cannot tell you
Correlation is backward looking. It describes 90 past days and says nothing about how two coins will move tomorrow. In a slump in particular, experience shows correlations rise, because at that point every position is being sold for the same reason. The protection a low correlation promises is weakest at precisely the moment it is needed.
There are four points we could not verify, and they belong to an honest picture. First, we calculated in US dollars; for you as an investor in the euro area, exchange rate movement comes on top and is not contained in these figures. Second, the measurement rests on daily values, so moves within a single day remain invisible. Third, we measured twelve coins and not the entire top 25; smaller coins outside that selection can behave quite differently. Fourth, a correlation says nothing about the size of the swings, because two coins can move in lockstep while fluctuating to completely different degrees.
The data source is an assumption too. Price data from an aggregator are blended prices across many trading venues. Recalculate using the prices on your own trading platform and the values will differ slightly. For the order of magnitude and for the ranking of the coins that makes no difference; for the second decimal place it does.
Checking correlation in your crypto portfolio: what to take away
The finding of this measurement: across 90 days the average pairwise correlation of the twelve largest cryptocurrencies is 0.61, across 30 days 0.64, and six of the eleven coins measured move with bitcoin at 0.77 or more. The number of positions in a portfolio therefore says little about actual diversification. Three steps follow from that, and the one-year clock stands above all three: every swap is a disposal for tax purposes and starts a new holding period, which is why the tax belongs in the balance against the risk benefit before any reallocation.
- List your weightings before you think about correlation. If two positions carry the bulk of portfolio value, their weighting decides more than any correlation figure. A portfolio tracker from our comparison shows the shares continuously and supplies the data you need for your tax return at the same time.
- Write down shared drivers beyond the price. Note which positions run on the same blockchain and which sit at the same place of custody. If you want to spread custody across several authorised trading venues, you will find the providers in our overview of crypto exchanges.
- For leveraged positions, calculate the distance for the whole account. Assume the same move for every position and check where the collateral stands then. Which platforms offer perpetual contracts and how their funding models differ is set out in our comparison of perp DEXs.
(As of September 21, 2026. This article is not investment advice. Prices and fee structures change; check the terms with the provider before you buy.)
Be the first to comment